`evorl.algorithms.erl.cemrl_td3.cemrl_td3_workflow`¶

Module Contents¶

Classes¶

CEMRLTD3WorkflowTemplate

A template for ERL workflow on TD3 Agent.

Functions¶

`build_cemrl_rl_update_fn`	K actors + 1 shared critic.
`cemrl_replace_td3_actor_params`
`create_dummy_td3_trainmetric`
`rollout_episode`

Data¶

DUMMY_TD3_TRAINMETRIC

API¶

class evorl.algorithms.erl.cemrl_td3.cemrl_td3_workflow.CEMRLTD3WorkflowTemplate(**kwargs)[source]¶

Bases: evorl.algorithms.erl.cemrl_workflow.CEMRLWorkflowBase

A template for ERL workflow on TD3 Agent.

evorl.algorithms.erl.cemrl_td3.cemrl_td3_workflow.DUMMY_TD3_TRAINMETRIC¶: ‘TD3TrainMetric(…)’

evorl.algorithms.erl.cemrl_td3.cemrl_td3_workflow.build_cemrl_rl_update_fn(agent: evorl.agent.Agent, optimizer: optax.GradientTransformation, config: omegaconf.DictConfig, agent_state_vmap_axes: evorl.agent.AgentState)[source]¶: K actors + 1 shared critic.

evorl.algorithms.erl.cemrl_td3.cemrl_td3_workflow.cemrl_replace_td3_actor_params(agent_state: evorl.agent.AgentState, pop_actor_params: evorl.algorithms.td3.TD3NetworkParams) → evorl.agent.AgentState[source]¶

evorl.algorithms.erl.cemrl_td3.cemrl_td3_workflow.create_dummy_td3_trainmetric(num: int) → evorl.algorithms.td3.TD3TrainMetric[source]¶

evorl.algorithms.erl.cemrl_td3.cemrl_td3_workflow.rollout_episode(agent_state: evorl.agent.AgentState, replay_buffer_state, key, *, collector, replay_buffer, agent_state_vmap_axes, num_episodes, num_agents)[source]¶